AI Writer

This repository contains two main pieces of the AI Writer project:

• AI_Writer_Client – a Next.js application that provides the user interface.
• AI_Writer_Service – a Cloudflare Worker service that hosts the API and business logic.

The client talks to the service via /api routes and the service relies on a single secret (SCREET_KEY) for authorization and other environment configuration.

---

📁 Repository layout

/ai-writer
  ├── AI_Writer_Client/      # Next.js frontend
  ├── AI_Writer_Service/     # Cloudflare Worker backend
  └── README.md              # (this file)

Both sub‑projects are independent; you can run them side‑by‑side or deploy them separately.

---

🔧 Prerequisites

• Node.js (v16+ recommended)
• npm or yarn
• Wrangler 3 (for Cloudflare Worker deployment)
• A Cloudflare account with Workers & KV namespace if you intend to deploy the service

---

🧩 Configuration & environment variables

Project	File	Variables	Notes
Client	.env.local	GEMINI_API_KEY	Gemini / LLM provider key used by the UI
Service	.dev.vars (local) or Worker secrets	SCREET_KEY	Primary secret used by getSecretKey()

Security: never commit secrets to git. .dev.vars is currently in the repo and should be added to .gitignore immediately; rotate SCREET_KEY and treat it as leaked.

To set a Worker secret with Wrangler:

wrangler secret put SCREET_KEY

Other configuration (e.g. database URL, KV names) lives in AI_Writer_Service/src/env.ts and the accompanying Typescript declarations; update as needed.

---

🚀 Running locally

Client

cd AI_Writer_Client
npm install
# add your Gemini key to .env.local
npm run dev

The UI will start on http://localhost:3000 by default and proxy /api requests to the Worker if configured accordingly (see next.config.ts).

Service

cd AI_Writer_Service
npm install
# create a local .dev.vars with SCREET_KEY and any other secrets
# example:
#   SCREET_KEY="your_dev_key"
npm run dev           # starts a local Miniflare server

The service will listen on http://localhost:8787 and exposes the same endpoints used by the client.

To run the database migrations (SQLite/D1) use the scripts in migrations/.

---

📦 Deployment

Client

Typical Next.js deployment (Vercel, Netlify, etc.). Make sure to set GEMINI_API_KEY in the host's environment.

Service

Use Wrangler to publish the Worker, e.g.:

cd AI_Writer_Service
wrangler deploy --env production

Configure KV namespaces and Worker secrets via the wrangler.toml/wrangler.json or through the dashboard.

Make sure to update the client’s next.config.ts API_URL to point at the published Worker.

---

🛡️ Secrets management

• Do not commit .dev.vars or `.env.local
• Rotate SCREET_KEY if it has been exposed.
• Consider using a vault or CI secret store for production values.
• Add a git pre‑commit hook or CI check (e.g. git-secrets) to prevent accidental leaks.

---

✅ Quick start checklist

1. Clone the repo
2. Install dependencies in both /AI_Writer_Client and /AI_Writer_Service
3. Create local environment files
4. Start the service with npm run dev inside AI_Writer_Service
5. Start the client with npm run dev inside AI_Writer_Client

---

📚 Further reading

• AI_Writer_Client/README.md – client‑specific instructions (UI, design tokens, etc.)
• AI_Writer_Service/AGENTS.md – architecture notes for the backend agents
• AI_Writer_Service/migrations – database schema history

---

🤝 Contributing

Feel free to open issues or pull requests. Ensure any new configuration is documented here and no secrets are checked in.

---

Generated by GitHub Copilot